Rotating detonation engines (RDEs) are emerging as a viable concept for high efficiency propulsion. RDEs use annular channels to support a continuously moving detonation wave that traverses the perimeter of the channel. Many other designs including radial and linear versions are possible. The main advantage of detonation-based combustion is the high wave speeds, which allow compact combustors to be used. Moreover, the pressure rise associated with the shock wave can be used to extract useful work. The APCL group develops high-fidelity computational solvers, analyzes physics of RDEs, and formulates reduced-order models for estimating performance characteristics.

Turbulent combustion is a complex physical process that involves strong coupling between chemistry, turbulent transport and fluid dynamics. Large eddy simulation (LES) combined with tabulated detailed chemistry provides alternative approach to accurately capture the transient effects of the turbulent flow and its interaction with the reaction process while maintaining an affordable computational cost. The essence of building a successful model using LES-based tabulated chemistry is to capture the dominant physics. While the dominant physics of conventional turbulent flames has been well studied (i.e. premixed and non-premixed flames are believed to be dominated by thermal and mass diffusion respectively), practical applications of turbulent combustion frequently involve other physical effects that are outside the “norm” of conventional configuration. For example, a turbulent flame flashback in heavy duty gas turbine represents a mixed-regime combustion between stratified/partially-premixed that further involves non-adiabatic effects. Another example is high altitude relight in aircraft combustors, which involves the transition of a flame kernel into a stabilized non-premixed flame (shown below). In these situations, 1) the choice of reduced-order model for tabulation (e.g. diffusion or premixed flamelet) needs to be made carefully, 2) secondary physical effects such as heat loss need to be modeled, and 3) extra modes of reaction need to be included into the combustion model, such as ignition under elevated enthalpy.

Environmental modeling continues to face challenges as a result of changing land-use practices, rapid urbanization, and growing awareness of environmental justice. The percentage of world population dwelling in cities continue to rise due to the increase in the number of cities, migration from rural to urban areas and transformation of rural settlements into urban area. Urban population density changes will cause cities to emit more GHGs. To counter this, the UN calls for cities and human settlements to be climate resilient and sustainable [1].
Recent studies have shown that around 90% of urban population is exposed to high levels of air pollution leading to health problems and terminal illnesses in certain cases [2]. This has prompted a growing interest in understanding topics such as air quality deterioration associated with pollutant dispersion, changes in wind flow patterns owing to physical setup and heat distribution variations as a result of land-usage. Previous studies on these topics have employed multiple tools, including CFD models which require hours of runtime for accurate results and predictive models which are not fully accurate but are fast. With factors like urban area morphology, meteorological data, and turbulent flows incorporated into simulation equations, understanding simpler topics like contaminant transport becomes more difficult. In addition to the physics aspect, detailed simulations of cities require considerable amount of computing resources. Therefore the main objective of the studies conducted by the APCL group is to accurately predict QOIs to help develop effective strategies to mitigate these adverse effects while simultaneously reducing the time-to-solution under a few minutes using a basic desktop device.